High throughput genotyping assays have allowed the staff of the Laboratory of Genomic Diversity (LGD) to accumulate a large amount of genotypic data on subjects who are members of six large AIDS epidemiology cohorts. The bulk of the genotypes are for diallelic single nucleotide polymorphisms (SNPs) that are within or closely linked to genes that have been nominated as AIDS candidate genes on the basis of what is known about their function. The LGD has previously reported significant associations between genotypes for candidate genes and HIV infection susceptibility or progression to AIDS outcomes. This work was largely done by analyzing data one SNP and statistical test at a time. In order to efficiently analyze a large amount of accumulating data, a set of programs is being developed to automate the analysis. The tests incorporated in the system include: 1) categorical analysis of infection status by genotype; 2) categorical analysis of disease progression by genotype; and 3) survival analysis using the Kaplan-Meier and Cox proportional hazard models.A first release of the analysis program has been produced for Apple Macintosh computers using the AppleScript language to create a graphical user interface and the SAS language to perform the statistical analyses. The automatic analysis allows multiple SNPs to be analyzed in a single execution of the program. The output is created in the form of a device-independent Postscript file that can be displayed on the computer screen or printed on any Postscript-compatible printer. The standard output consists of an 11 page report for each SNP with results displayed either as tables or graphs. Automation of AIDS Candidate Gene Analysis